Project Details


This award is part of the NSF effort to promote significant advances in the fundamental understanding of cancer biology made possible through multidisciplinary research that involves experts in theoretical physics, applied mathematics, and computer science.

Achieving durable control of metastatic solid tumors will require high-order targeted therapeutic combinations, because single-agent therapeutics eventually become thwarted by the development of tumor drug resistance. However, design of combinatorial regimens cannot be done by empirical trial and error in the clinical setting. The goal of the project is to blend a systems biology network-based theoretical framework with an integrated experimental and analytical program in order to address the combinatorial regimen challenge in oncology. Based on areas of exemplary clinical need, investigator expertise, and the availability of patient-derived tumor tissue, the project will focus on BRAF-mutant melanoma and PIK3CA-mutant, estrogen receptor positive (ER+) breast cancer as initial tumor types in which to pilot the approach. In addition the project will offer interdisciplinary training and research experience to postdoctoral and clinical fellows, graduate students, and indirectly to all members of the groups who participate. Professional development of all trainees will be enhanced by yearly meetings of the whole project team which will include tutorials on modeling and experimental methodologies. A symposium on the quantitative science of cancer will be organized at the Dana Farber Cancer Institute during the third year of this project. Team members are also committed to broadening the participation of women and under-represented minorities in STEM fields by pro-active recruitment and mentoring.

The project will integrate dynamic modeling of signal transduction pathways relevant to cell proliferation and apoptosis, genomic and evolutionary analyses of tumor cells, and systematic cell death and therapeutic resistance studies. The dynamic models will be informed, tested, and iterated using experimental approaches applied to relevant cancer model systems. The experiments leverage emerging technologies such as pooled genome-wide open reading frame screens, dynamic BH3 profiling of cancer cells' closeness to the apoptotic threshold, whole exome sequencing and single cell RNA-seq analysis. The models will recapitulate steady state signaling network activation, acute adaptive effects of treatment (e.g., feedback dysregulation) and the range of drug-resistant states that may emerge following longer-term drug exposure. Tumor cell heterogeneity will be represented by the implementation of different initial configurations or state overrides of network components. Using newly developed systems control methodologies, the models will be used to prioritize drug combinations and dosing/scheduling principles for in vitro and in vivo testing. The final result will be a theoretical and experimentally validated approach that can be generalized across many other cancer types. This project develops a new framework to address cancer as a deregulated complex dynamical system and it will lead to an improved understanding of adaptive and acquired drug resistance mechanisms. The project will make a significant contribution toward a major goal of cancer precision medicine, namely the identification of optimal high-order combinations for individual cancer patients. The project will also establish new connections between evolutionary theory and dynamical systems theory. The theoretical and methodological advances will be applicable or adaptable to other cancers and diseases in general, leading to potentially transformative impacts on human health.

This proposal is cofunded by the Physics of Living Systems Program in the Physics Division and the Systems and Synthetic Biology Program in the Molecular and Cellular Biosciences Division.

Effective start/end date6/15/1611/30/20


  • National Science Foundation: $240,480.00


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.